Unlike most of the NAEP operational studies, the population target for the long-term trend studies is defined by age rather than grade. The three age populations for the 2004 study are as follows:
The sample was designed to achieve a nationally representative sample of each of these age populations. The target population included all students in these age ranges enrolled in public and private schools. The samples were selected based on a three-stage sample design:
The second‑stage samples of schools were selected with probability proportional to a measure of size based on the estimated number of age-eligible students in the school. This, in turn, was estimated by applying population-level percentages of age-eligible students within each grade to estimated grade enrollments for each grade, and aggregating to an age-eligible total for the school.
Each selected school was asked to provide a roster of all of its students who were age-eligible. A fixed sample of students was selected from these rosters (128 for nine- and thirteen-year-olds, 121 for seventeen-year-olds), excepting smaller schools with total age-eligible counts less than the sample size. The combination of using the number of age-eligible students as a measure of size and a fixed sample size of students is intended to give a self-weighting sample (each student has an equal probability of selection). In practice, differences between the estimated age-eligible enrollment for the school and the actual size of the finalized roster introduce slightly differing student probabilities across schools.